(vii) The Korean authorities considered the extent to which other factors were causing injury to the domestic industry

4.585 The Korean authorities evaluated the following other factors:

(a) The milk quality dispute

4.586 The Korean authorities evaluated the arguments presented by interested parties, including the Korea Food Industry Association ("KFIA") and the Korea Dairy Cow Breeding Association ("Breeding Association"). ²⁷⁴ The KFIA contended that the milk quality dispute contributed 87.17 per cent to the rise in the milk powder inventory and the increase in SMPP contributed 4.80 per cent. ²⁷⁵ The Breeding Association contended that increased imports of SMPP contributed 61.9 per cent to the increased milk powder inventory and the milk quality dispute contributed 17.0 per cent. ²⁷⁶

4.587 The Korean authorities rejected the analysis of the KFIA based on flaws in its data and analysis. ²⁷⁷ The authorities also rejected the Breeding Association's analysis because it failed to submit basic materials to support its claims. ²⁷⁸ Instead, the authorities developed three statistical and econometric models for estimating the volume of milk production for the period covering November 1995 to June 1996, assuming the quality dispute had not occurred. ²⁷⁹ The authorities concluded that "[b]ased on the results of three models, the figures reflecting the difference between estimated and actual volume of milk production indicate that the quality dispute no longer had effect as of January 1996. ²⁸⁰" As part of the OAI's investigation to determine causality, the OAI also employed a commonly-used econometric technique called the "Granger causality test".The test results, checking up to six lags in variables to maximize accuracy, indicated that the increased imports of SMPP caused serious injury to the domestic industry. ²⁸¹ Taken together, the Korean authorities determined that while the milk quality dispute had an effect for a few months, the imports of SMPP had a negative effect on the domestic industry throughout the entire investigation period.

(b) Influence of reduced imports of milk powder

4.588 The Korean authorities evaluated the influence of reduced imports of milk powder in its causation analysis. The Korean authorities essentially determined that:

(c) Review of demand

4.589 The Korean authorities also examined other factors potentially causing serious injury to the domestic industry relating to demand for domestic raw milk and milk powder. With respect to consumption, they stated that "given the overall increase in consumption, the injury to the domestic industry is not attributable to a decrease in consumption." ²⁸³ With respect to whether the shift to flavoured milk from white milk was based on changing consumer preferences, the authorities found that the shift was "attributable to milk processors seeking to change the production structure in order to maximize their profits through the use of cheaper imported products, as opposed to changes in consumer preference." ²⁸⁴ With respect to consumption of final dairy products, they determined that dairy products consumption increased during the period from 1993 to June 1996, and "[t]herefore, it was not regarded as a cause of injury to the domestic industry." ²⁸⁵

4.590 With respect to the effect of the situation in other dairy products sectors, the Korean authorities examined the white milk, milk powder preparation, evaporated milk, butter, and cheese sectors. Because the production and consumption of these dairy products increased during the period of the survey and because their use of domestic raw milk and domestic milk powder also increased, the authorities found that "no connection could be established between the effect on these dairy products and injury to [the] domestic industry." ²⁸⁶

(d) Review of price decision by the government

4.591 Finally, the Korean authorities evaluated any potential injury caused by the suggested government price for raw milk. The authorities found that the suggestion of such a price did not adversely affect the domestic industry because it was not obligatory and was intended as a suggested reference price for contracts between dairy households and milk processing firms. ²⁸⁷

(f) Additional arguments by the European Communities made at the second meeting of the Panel with the parties

4.592 At the second meeting of the panel with the parties, the European Communities further advanced their arguments under Article 4.2(b) as follows:

4.593 The European Communities point to the flaws in the conclusions drawn from by Korea from its econometric models. The models do show a shortfall in consumption continuing until June 1996 but it is then stated at pages 68 to 70 of the OAI Report that this cannot be due to the milk quality dispute as this ended in February. No other explanation is advanced for the shortfall. This is perfectly circular reasoning.

4.594 Korea alleges it performed a "Granger causality test", without in any way explaining where it is to be found in the OAI Report, how this test works, precisely what its results demonstrated, what variables were used in relation to "checking lags", what level of statistical accuracy was demonstrated in the test results, etc. This test has never been mentioned before by Korea and cannot be taken seriously without considerably more information being given as to its nature and its application to this situation.Indeed, the European Communities asked, what happens when it is applied to the milk quality scandal?

(g) Additional arguments by Korea made at the second meeting of the Panel with the parties

4.595 At the second meeting of the panel with the parties, Korea further advanced its arguments under Article 4.2(b) as follows:

4.596 The European Communities argue that the econometric models used by the OAI were flawed in that they examined production, instead of consumption.Korea notes that:

4.597 In response to a question by the Panel ²⁸⁸, Korea explained that in order to establish any causal relationship between the imports of SMPP and the increase in inventory, the OAI used the econometric technique referred to as "Grainger Causality" test. The following parameters or techniques were applied:

4.598 Therefore the calculated ADF statistics showed that the import of SMPP and inventory of milk powder are integrated of order one, like most other macroeconomic time series. As the first differenced variables were revealed to be stationary, they were used to obtain the result of causality tests. The most commonly used causality test is the Granger test, and this was used to obtain the result. Pairwise Granger causality tests, checking lags up to 6, revealed that imports of SMPP caused an increased in inventory of milk powder at 5 per cent in four out of six cases and at 10 per cent in one case. The causation was negated at 10 per cent only in one out of six cases. The calculated F statistics were the following:

4.599 Therefore, on the basis of the above, the OAI concluded that the econometric causality tests revealed a causal relationship running from the import of SMPP to the increase in inventory of milk powder.

4.600 In response to a question by the Panel ²⁸⁹, Korea clarified the substitutability of raw milk, milk powder and SMPP. As stated in Section II.2 of the OAI Report, SMPP can substitute domestically produced raw milk and milk powder to produce flavoured milk, fermented milk, ice cream and cookies.Based on the import clearance document submitted by foreign exporters, the Korean authorities verified that SMPP can be used to produce milk based beverages, ice cream and cookies.That the imported SMPP and domestic milk powder are in directly competitive relationship is evidenced by the fact that the primary difference between the two products is that the former contains 75 per cent to 85 per cent milk powder and 15 per cent to 25 per cent whey or malt concentrate. ²⁹⁰ Moreover, for commercial purposes, the imported SMPP and domestic products share common end uses.Imported SMPP and domestic products can be used to produce flavoured milk, fermented milk, ice cream and cookies. ²⁹¹ As a further indication that the imported SMPP and domestic products are directly competitive products, the Korean authorities found that, of the total purchase of basic materials, five dairy processing companies increased their purchases of SMPP from 3 per cent in 1993 to 23.8 per cent in 1994, 29 per cent in 1995 and 53.3 per cent in January-April 1996. ²⁹²

4.601 In producing flavoured milk and cookies, one unit (1 kilogram) of SMPP can substitute one unit(1 kilogram) of milk powder or ten units of raw milk.In producing ice cream and fermented milk, one unit of domestic milk powder can be replaced by anywhere between 0.8 to 1.2 units of SMPP, depending on the dairy processing company. The Korean authorities decided that the use of one to one substitution rate between imported SMPP and domestic milk powder to produce ice cream and fermented milk was appropriate because, based on the data collected, the usage rate would average out to one unit to one unit rate.

4.602 The Korean authorities were aware that (1) the domestic products and the imported SMPP had physical differences but shared common end uses and (2) substitutability between the imported SMPP and domestic milk powder was on a one unit to one unit basis and that ten units of raw milk are needed to produce one unit of milk powder. The conversion rate of ten raw milk units to one milk powder unit is reflected in various places of the OAI Report. For instance, charts in Sections III.5.A and VI.2.A reflect this conversion rate. The substitution rate of one unit to one unit between SMPP and milk powder is reflected in the Note at the bottom of the chart in Section III.5.A which states "for computation of demand and self sufficiency rates, the amounts of SMPP and imported milk powder calculated in terms of domestically produced raw milk were used." Although this is not the most eloquent translation, it indicates that the Korean authorities were familiar with the fact that ten units of domestic raw milk were equivalent to one unit of SMPP or one unit of imported milk powder.

H. Claims under Article 5.1 of the Agreement on safeguards

(a) Claim by the European Communities

4.603 The European Communities claim that Korea violated Article 5.1 of the Agreement on Safeguards by failing to show that the measure was necessary to prevent or remedy serious injury and to facilitate adjustment; by failing to demonstrate that a quota was the most suitable to prevent or remedy serious injury and to facilitate adjustment; and by imposing a quota lower than the average of imports in the last representative three-year period preceding the application of the measure for which statistics were available.The following are the EC arguments in support of these claims:

4.604 Article 5.1 of the Agreement on Safeguards provides that :

4.605 The fact that safeguard measures are "limitative and deprivational in character or tenor and impact upon Member Countries and their rights and privileges and upon private persons and their acts" was clearly recognized by the Appellate Body in its report in US - Underwear. ²⁹³ In the light of that characterization, the Appellate Body drew the conclusion that an importing Member should not be allowed "an enhanced ability to restrict the entry into its territory of goods in the exportation of which no unfair trade such as dumping or fraud or deception of origin is alleged to be proven" by taking safeguard action beyond the strict limits laid down in the relevant WTO provisions, which would result in "excluding more goods from the territory of the importing Member". ²⁹⁴

4.606 Some of the limits built in the WTO safeguard measures regime relate to the measures themselves, notably to their scope, level and type. Those limits are laid down in Article 5.1. Besides recalling that safeguard measures must be necessary to remedy serious injury, as provided for in Article XIX:1(a), ²⁹⁵ Article 5.1 of the Agreement on Safeguards further requires that the temporary protection from foreign competition must be necessary to facilitate adjustment. ²⁹⁶ The rationale for this provision is clearly that protection of an inefficient industry sector with no recovery prospects by means of safeguard measures should be excluded.

4.607 It follows from the foregoing that a Member of the WTO seeking to take a measure under the Agreement on Safeguards must show that such a measure is, in its scope and level, necessary to remedy the injury suffered by the domestic industry and necessary to facilitate its adjustment. In this respect the European Communities note that Korea failed to provide any justification as to the reasons why the quantitative restrictions applied were necessary to remedy the alleged serious injury and to facilitate adjustment. In particular, Korea did not submit any information as to adjustment plans to restore the domestic industry's competitiveness while temporarily shielding it from foreign competition. On the contrary, Korea took safeguard action in the context of a protected market. It is clear to the European Communities by omitting to give any consideration to adjustment plans, a fortiori Korea has failed to examine how that measure could be necessary, or even helpful, to their implementation.

4.608 For these reasons, the European Communities submit that Korea violated its obligations under Article 5.1, first sentence of the Agreement on Safeguards.

(i) The most suitable measures to remedy serious injury and facilitate adjustment

4.609 Article 5.1, last sentence, of the Agreement on Safeguards, further building upon the remedial character and the adjustment objective of safeguard measures, also requires Members to "choose measures most suitable for the achievement of these objectives." The ordinary meaning of this clause already suggests that both the objective to remedy the injury and the objective to facilitate adjustment also limit the type of measure which a Member may adopt, in addition to its level or scope. This construction is further reinforced in the light of the principle of effective interpretation of treaties, based on which where a treaty provision can be subject to several possible interpretations, preference should be accorded to the one giving that provision its effect. ²⁹⁷ In fact, if the second sentence of Article 5.1 were also to be interpreted as a limit to the level or scope of a measure, rather than as including some additional element, it would be redundant in the light of the first sentence.

4.610 The conclusion that the choice of the type of safeguard measure is limited under Article 5.1 of the Agreement on Safeguards is confirmed if that provision is interpreted in its context. In fact the Agreement on Safeguards includes a partially different regime for quotas and tariff measures. In the first place, a maximum limit to the level of protection, additional to the injury level and the adjustment required, is imposed by Article 5.1, second sentence, providing that "if a quantitative restriction is used, such a measure shall not reduce the quantity of imports below the level of a recent period which shall be the average of imports in the last three representative years for which statistics are available." Second, in case of provisional measures Article 6 of the Agreement on Safeguards limits the possibility of action to tariff measures only. Also, the different nature and impact of the various protection measures has been expressly recognized in the regulatory framework established by the WTO Agreement and its annexes. Specifically, Members recognized that "price-based measures", that is, "measures with an impact on the price of imported goods" "have the least disruptive effect on trade". ²⁹⁸

4.611 It follows from the foregoing that if a WTO Member seeks to take a safeguard measure, the choice of the type of measure also needs to be justified in terms of its adequacy to remedy injury and facilitate adjustment. The European Communities therefore conclude that, by failing to consider whether other types of measure than a quota would be the most suitable to remedy serious injury and to facilitate adjustment, Korea further violated its obligations under Article 5.1 of the Agreement on Safeguards.

(ii) Quota lower than average of imports in the last representative three-year period preceding the application of the measures for which statistics were available

4.612 Article 5.1, second sentence, of the Agreement on Safeguards provides that

4.613 The average of imports into Korea for July 1993-June 1996 was lower than the average for the period January 1994-December 1996. The European Communities submit that, by calculating the quota level on the basis of import data relating to the period July 1993-June 1996, rather than to the period January 1994-December 1996, the Korean authorities violated Article 5.1 of the Agreement on Safeguards, second sentence, since July 1993 - June 1996 data did not relate to the "last three representative years for which statistics are available" and the resulting quota was lower than allowed by that provision, without justification being provided.

4.614 In order to demonstrate this claim the European Communities first examine the meaning of Article 5.1, second sentence of the Agreement on Safeguards, and then turns to the three years which are relevant under that provision.

To continue with Article 5.1 of the Agreement on Safeguards

²⁷⁴ OAI Report at 67-68. See, also Notification at V.2.1.

²⁷⁵ OAI Report at 67. See, also Notification at V.2.1.a.

²⁷⁶ OAI Report at 67.

²⁷⁷ See, OAI Report at 68; Notification at V.2.1.a.

²⁷⁸ See, OAI Report at 68.

²⁷⁹ Id. In its First Submission, the EC contends that the competent authorities "fail[ed] to make any analysis" of the effects of the milk quality dispute. In its Oral Statement, the EC states that "[n]o detail is given of the analyses which led to this remarkable conclusion" and that the milk quality dispute was "lightly dismissed." Korea again suggests that the EC review the OAI Report of which it has had a copy in English at least since early 1997.

²⁸⁰ OAI Report at 69.

²⁸¹ The Granger causality test indicated for example, that the increased imports of SMPP caused an increase in milk powder inventory at 5 per cent level of significance. In hypothesis testing 5 per cent level of significance means that the level of confidence of the test result is 95 per cent (see, for example, Damodar N. Gujarati, Basic Econometrics, 2nd. Edition, McGraw Hill: NY, 1988, p. 99).

²⁸² Notification at V.2.2.

²⁸³ Notification at V.2.3.1.

²⁸⁴ Notification at V.2.3.2. In its Oral Statement, the EC disagrees with the competent authorities' judgement on this issue. Korea considers that the EC should not be allowed to assume the role of the investigating authority.

²⁸⁵ Notification at V.2.3.3.

²⁸⁶ Notification at V.2.3.6.

²⁸⁷ See, Notification at V.2.4.

²⁸⁸ The Panel recalls that the question was: "How was the 'Granger' test applied in this case?"

²⁸⁹ The Panel recalls that the question was: "In Korea's first submission, it is stated that SMPP products are competing directly with domestically-produced raw milk and pure milk powders. Could Korea elaborate? Does this mean that 1 unit (say 1 Kg) of SMPP can substitute 1 unit of milk powder and/or 1 unit of raw milk for the production of downstream products? Please provide the Panel with exact figures for each of the downstream product (e.g., yoghurt, flavoured milk, ice-cream, cheese etc...). Was the Korean Trade Commission aware of these facts at the time of its investigation? Did it take them into account in its report and if so how, and where is it indicated in the OAI Report?

²⁹⁰ See, OAI Report at Section II.4

²⁹¹ See, OAI Report at Section IV.1.B.(4).(a)

²⁹² See, OAI Report at Section IV.1.B.(4).(B).(i)

²⁹³ See, Appellate Body report in US - Underwear, 10 February 1997, p. 9.

²⁹⁴ Id.

²⁹⁵ Article XIX:1(a) refers, in virtually identical terms, to the "extent and time necessary to remedy the injury".

²⁹⁶ See, also the Preamble of the Agreement on Safeguards, second to last paragraph, "[r]ecognizing the importance of structural adjustment and the need to enhance rather than limit competition in international markets".

²⁹⁷ The principle of effectiveness in treaty interpretation was recognized as an appropriate interpreting principle by the Appellate Body in US - Underwear, p. 5.

²⁹⁸ See, the Understanding on the Balance-of-Payments Provisions of the General Agreement on Tariffs and Trade 1994, para. 2. Para. 3 is even more explicit in providing that "Members shall seek to avoid the imposition of new quantitative restrictions for balance-of-payments purposes unless, because of a critical balance of payments situation, price-based measures cannot arrest a sharp deterioration (...). In those cases in which a Member applies quantitative restrictions, it shall provide justifications as to the reasons why price-based measures are not an adequate instrument to deal with the balance-of-payments situation."